Today’s Dad Joke
Just so everyone’s clear, I’m going to put my glasses on.
Housekeeping
- C-Span Archives Undergrad Research Competition
- bit.ly/ccseurc20
Main questions
- How did you play differently in different versions?
- What in real life is similar to each version of the game?
- How did you decide whether or not to help others?
- What does this have to do with social networks and social capital?!
Debrief
- What did you think about while playing?
- How did you feel as you played? Did those feeling change across different versions of the game?
- When do you help others in real life?
Debrief
- Were resources and abilities distributed fairly?
- How should resources be distributed in the real world?
- What role did communication play? How did your group make decisions?
- Did you work to make things equal? Why or why not?
- What kinds of inequality appeared in the game?
Debrief
- Did anyone sanction someone else? Why?
- How did others respond to being sanctioned?
- What is similar to a “sanction” in the real world?
Aspects of visualizations
Nodes
- Information can be conveyed by:
- Shape
- Typically categorical (e.g., gender, age range)
- Size
- Often a network measure, but can be something about node
- Color
- Often community detection
Examples

With node shapes

Nodes sized by betweenness centrality

Edges
- Size
- Typically represent weight of relationship
- Color
- Typically represents different types of relationships
Edge width as weight

Edge color as type

Position
- Can represent literal distance
- Or social distance
- Formal hierarchy
- Degree centrality
Today’s activity
- Start working on DataCamp Network Analysis, chapters 2 and 3
Today’s Dad Joke
Within minutes, the detective knew exactly what the murder weapon was.
It was a brief case.
Housekeeping
- Self-assessment overview
- R
- Coming with questions / Office hours
- Readings
- Too long
- Put on Brightspace
- Posting slides
- In class
- Break
- Dealing with confusion
- More group discussion
- Questions in class
- Writing on projector vs. on board?
DataCamp Review
- neighbors()
- ego()
- make_ego_network()
- Come back to:
- lapply
- randomization
- transitivity
- cliques
What makes for a good network visualization?
Kozo Sugiyama’s rules:
- Lines should be straight.
- Lines should be far apart from one another.
- Lines should not cross or touch.
- Lines should be easy to follow from one node to another.
- Nodes that connect should be close.
- Nodes that are most central should be close to the center of the graph.
- Nodes that are similar in some way should be placed near one another.
Activity
Find and Assess a network visualization
- Use Google Images (or similar) to find an image of a network graphs.
- What do you think the image does well and poorly? Does it follow Sugiyama’s rules?
- Put the image (and your assessment) on the padlet at https://padlet.com/jdfoote1/networks (link is also on wiki)
Discussion
- Did you agree with Sugiyama’s rules?
- What types of color/shape/size approaches were effective?
Visualization Competition
- Dutch school dataset (on wiki)
- In groups of 2-3
- Come up with question
- Come up with visualization approach
- We will work on this more next Thursday
Homework
- Scott Feld guest lecture
- Summary and questions assignment